Journal of Conference Abstracts

Evolution of the Mid-Atlantic Ridge Between 28° and 31°30'N During the Last 10 Ma: Implications from 3D Gravimetric and Magnetic Study

C. Rommevaux CNRS, Lab. de Pétrologie, Université de Paris 6, 4 Place Jussieu, 75252 Paris Cédex 05, France

cer@ccr.jussieu.fr

J. Pariso 54 rue Peyrolieres, 31000 Toulouse, France

Ph. Patriat CNRS, Equipe Cinématique et Dorsales, Institut de Physique du Globe, 4 Place Jussieu,

75252 Paris Cédex 05, France

J.-C. Sempéré School of Oceanography, Univ. of Washington, Seattle, WA 98195, USA

C. Deplus CNRS, Lab. de Gravimétrie et Géodynamique, Institut de Physique du Globe, 4 Place Jussieu,

75252 Paris Cédex 05, France

Multibeam bathymetry (SeaBeam and Hydrosweep), gravity and magnetism data were collected along the Mid-Atlantic Ridge (MAR) between 28° and 31°30'N, during two high resolution surveys (SARA - RV J. Charcot, May 1990; EW9210 - RV M. Ewing, Oct. 1992). These surveys were conducted to study the evolution of axial segmentation with time. Data coverage extended up to 10 Ma (anomaly A5) on either flank. In this area, seven morphological axial segments were identified, separated by right-stepping offsets including the Atlantis transform fault (30°05'N, ~70 km) and a series of non-transform discontinuities (NTD). The morphology of the spreading centers differs on either side of the Atlantis Fracture Zone (FZ). South of the transform, the rift valley is well defined, and the bathymetric expression of the NTD's over the flanks consists of a series of aligned basins forming V-shaped troughs pointing south. North of the transform, the spreading axis consists of en échelon volcanic ridges located on a rift valley oblique to spreading direction. The bathymetric expression of the discontinuities changes at 4 Ma, with well defined transform faults parallel to the spreading direction between 10 and 4 Ma, and NTDs with small offset from 4 Ma to the present axis. A 3D gravimetric and magnetic analysis was conducted in order to constrain the processes responsible for this difference on the observed morphology, and to characterize the structure of the MAR lithosphere.

Low gravity anomalies over the MAR axis are commonly interpreted as due to along axis variations in crustal thickness within segments resulting from focused mantle upwelling. In our area, the residual gravity anomaly exhibits substantial variations along and across the ridge axis, which can be accounted for by variations in crustal thickness of up to 3 km. For the segments south of the Atlantis FZ, the segment mid-points and the segment discontinuities have been associated respectively with thick crust and thin crust for the last 10 Ma, suggesting the segments act as long-term individual units in terms of melt delivery. In contrast, for the segments north of the fracture zone, the crustal thickness pattern is very complex, with a change around 4 Ma. The crustal thickness appears thicker from 4 Ma to the present axis and not really correlated with the morphological segmentation, compared to the 4 - 10 Ma period. This result suggests that mantle upwelling became more robust recently, north of the fracture zone. In all of the paleo-segments in our study area, variations in computed crustal thickness are observed across-axis, but are not correlated between adjacent spreading segments, suggesting temporal and spatial variations in melt production.

The magnetic field data show that, in addition to the expected reversal chronology, there are important spatial variations in the magnetic field observed over our study area, which are directly related to the paleo-segmentation pattern. To understand the source of these anomalies, we performed a 3D inversion of the magnetic field (taking into account the bathymetric variations) and calculated a magnetization distribution. The magnetization exhibits a sharp decay from the axis to older lithosphere which could be attributed to progressive low temperature oxidation of basalt. In ~10 Ma old crust, we observe an abrupt increase in magnetization relative to more magnetic material or to a thicker magnetic source layer. A two-layer model, consisting of a shallow extrusive layer and a deeper intrusive layer with sloping polarity boundaries, can account for this increase at anomaly 5 time (a reversal epoch of relative long duration). In crust older than ~2 Ma, the discontinuity traces are characterized by low magnetic anomaly amplitudes, but high, positive magnetization. In addition, within these traces the reversal pattern is missing. This observation could be due to a high induced component of magnetization, produced by serpentinized lower crust/upper mantle, and masking the contribution of basalts to the magnetic anomaly signal. South of the Atlantis FZ, high magnetization is correlated with bathymetric troughs at segment end points and lower magnetization is associated with bathymetric highs at segment midpoints. North of the fracture zone, there are no systematic along-axis variations in crustal magnetization, and off-axis NTD traces are not associated with high magnetization values. This difference suggests that these variations are a result of the nature of crustal formation process, with more focused mantle upwelling to the south of the Atlantis FZ than to the north.

The differences in morphology, crustal thickness and magnetization pattern on either side of the Atlantis FZ suggests different processes of crustal production. To the south, focused mantle upwelling has been maintained for at least 10 Ma. The orientation of bathymetric troughs shows a migration of the NTDs suggesting relative/absolute motion changes or local variations of amount of melt delivered to spreading segments. To the north, a change of melting delivery appears around 4 Ma: before this time, low crustal production and focused mantle upwelling; after this time, a robust mantle upwelling relatively continuous along the axis. This difference in the crustal production may be related to a change in the spreading direction, which occurred ~7 Ma ago and resulted in modifications in the geometry of the plate boundary.